Improving Access to High-Quality Mental Health Care for Veterans: Proceedings of a Workshop (2023)
Chapter: 3 Leveraging VA's Status as a National Health Care System
VA and its system of health care coverage for veterans is unique in the United States because it is a national program. This chapter summarizes speakers’ discussions of other national health care systems that have notable mental health components and also highlights two U.S. delivery systems whose approaches VA could potentially learn from.
INTERNATIONAL EXEMPLARS
Asaf Bitton, executive director of Ariadne Labs in Boston, described his organization as an academic implementation science center led by clinicians, with a mission to design, test, and spread scalable solutions to narrow gaps between known best practices and what is done in routine clinical practice. “Health care is full of know-do gaps,” he said, [which are] “things we know we should do but that [don’t] routinely happen.” Bitton said that it’s usually not one person’s fault but instead because of a system that was not designed to be responsive and meet the needs of people it is trying to serve. These gaps result often in preventable suffering, and narrowing or closing them can improve health, reduce suffering, and increase overall system performance. Although even health systems that work well may have flaws, they always look to see if they can generate ideas to help solve problems. Bitton reviewed three different health systems and the models they use: Increasing Access to Psychological Therapies (IAPT) in the United Kingdom, Trieste in Italy, and the Nuka/Southcentral Foundation in Alaska.
Even though the UK National Health System is unique, it shares many similarities with the VA system insofar as both are national, integrated, and comprehensive, and they integrate financing and care delivery. Bitton stated that after large evidence reviews in the early 2000s that showed that UK health care services were not meeting the population’s needs (McManus et al., 2009), the government began to build a national model that included countrywide standards, training, and data collection. Bitton explained that it also created a hub-and-spoke model, with a central coordinating facility and multiple entry points into the system. The model, which continues today, divides the mental health workforce into two groups. The first includes high-intensity therapists with more hours of training. The second includes lower-intensity psychological well-being practitioners who focus more on community engagement. Before IAPT, Bitton said that less than 40 percent of UK patients with depression and anxiety received any evaluation, and less than half were even known to the system as having these challenges. To address this gap, a national measurement system was built based on PHQ-9 and GAD-7 measurement tools and incorporated patient-reported outcome measures around mental health. Overall, he summarized, the system made a major investment in data collection, standardized training, and a stronger, better organized workforce (Clark, 2018,
2019). In the first 3 years, IAPT saw a 45 percent full recovery rate, resulting in a savings of £272 million through reduced health care use. Therapists were able to reach 1.2 million people in 2021, showing the tremendous reach of the program, which was renamed as “NHS Talking Therapies” in 2023 (NHS England, n.d.).
The second model is from the northern Italian city of Trieste, centered on people with serious mental illness; it began in 1971, during a time of deinstitutionalization. It focused on ensuring community support through a series of touchpoints and multidisciplinary staffing, prioritizing integrating patients back into communities, the workforce, and greater society. The community was accountable for overall outcomes, so it would track patients and note the services they needed. Workforce training was standard for psychiatrists and psychologists but emphasized social workers and partnerships with volunteer and social promotion associations. The model also works with social cooperatives to train and employ people with mental health and substance use disorders, creating a social lattice for integration. Hundreds of businesses and community members volunteer to employ people from this program, Bitton added, knowing that it may take extra care and attention to help them on their journey to recovery. The model focused more on social than medical indicators for measurement and spent more time looking at outcomes than medical processes. For example, outcome measures included the proportion of people successfully living independently with or without support and levels of engagement of community members and associations. Between 1971 and 2000, psychopharmaceutical spending reduced by half, with a 45 percent regional reduction in suicide rates. Only 18 out of 5,000 people treated by Trieste’s mental health department are involuntarily housed, but Bitton acknowledged large variations in uptake nationwide. The northern region has adopted the model much more successfully than Southern Italy, largely due to political and funding challenges.
Last, Bitton reviewed the Southcentral Foundation’s Nuka system, serving about 60,000 Alaska Natives. The model is centered on empanelment to primary care teams, high levels of patient access to them, and integrated team-based behavioral health care. The program leans on an integrated and comprehensive care team consisting of behavioral health consultants and trained community health workers (CHWs) and focuses on easy access to care, open scheduling, expanded office hours, home visits, and team care. Training is based on a commitment to the community and customer-service focus. He explained that it does not call people “patients” but instead thinks of them as “customer–owners” of the system. In addition to training behavioral health consultants and CHWs, the program also trains family wellness warriors in trauma-informed care and leverages group therapy through learning circles, based on Alaska Native storytelling and sharing. Using these people-centered
approaches and extensive measures of patient experience and process measures, Bitton noted huge decreases over the past 20 years in emergency room visits; substance use visits; untreated depression; and even indirect important social outcomes, with a 20 percent documented reduction in unintentional injury visits to the emergency department, suggesting widespread downstream effects (see Figure 3-1).
In summary, Bitton said all three models he presented focus on standardized training, data collection, use of data for meaningful whole-health community engagement, and work to demedicalize and broaden available solutions.
IMPROVING MENTAL HEALTH CARE QUALITY IN THE UNITED STATES: INTERMOUNTAIN HEALTH CARE
Brenda Reiss-Brennan, chief clinical science officer at Intermountain Health Care, offered three concepts to potentially guide VA’s forward progress: complexity, relational reciprocity, and matching. In describing Intermountain, Reiss-Brennan said it is an ecosystem that has been working to scale relational
NOTE: EC = emergency care; RPMS = resource and patient management systems.
SOURCES: Asaf Bitton presentation, April 20, 2023; David et al., 2013.
team networks around an organized common goal for everyone, while also having a common language for mental health. Part of this process included shifting terminology from “mental illness” to “mental health and wellness.” In the early stages of this effort, it built infrastructure to integrate mental health across the entire medical experience instead of adding mental and behavioral health as a separate program. Talking with and listening to its primary care clinicians, staff, and patients, Intermountain realized the complex needs of the primary care population, ranging from domestic violence and depression to job loss, chronic pain, and substance use. It used implementation and complexity science to build an ecosystem with a clinical and operational team-relational process incorporating mental health at every contact, which revealed five elements that are foundational to its mental health integration program: leadership and culture, clinical workflow, information systems, financing and operations, and community resources (with the patient and family at the center) (see Figure 3-2).
Reiss-Brennan emphasized that it required a large network of teams across professions and practice expertise to make this system successful. She highlighted that the clinic receptionist had the most influential change input to transforming primary care to align with what patients and their families truly needed. Because of this, receptionists are now key members included in the leadership team for building and maintaining the program. For mental health integration to succeed, Reiss-Brennan continued, it requires matching
SOURCE: Brenda Reiss-Brennan presentation, April 20, 2023.
combinations of connected team resources to the complexity of patient and family needs. The integration program was able to stratify everyone coming into primary care to access mental health care through a trained integrated team based on their level of physical, social, and mental health complexity. Reiss-Brennan said that Intermountain built an ecosystem to support a variety of needs and has facilitated effective communication and decision-making as well as operational and social connectedness. The program built a longitudinal registry to support and monitor team performance and follow patients over time to accurately track and report cost and quality outcomes. This outcome information can be useful for operational and clinical teams to engage and sustain the hard work of implementation. More than 700,000 people have been tracked in Intermountain’s depression registry, which has been integrated with other chronic disease registries since 2000.
Intermountain also conducted ethnographic interviews on the lived experience of patients, families, and staff to measure the success of these new systems and programs. Responses were positive and acknowledged that provider teams were communicating and patients themselves were connected to the team (Reiss-Brennan, 2014). In 2016, Intermountain looked back over 10 years (2003–2013) across more than 100,000 participants and found that its integration of mental and physical health through primary care teams resulted in better clinical outcomes and lower costs (Reiss-Brennan et al., 2016). Emergency department visits, hospital admissions, primary care encounters, and payments all reduced. Although this pointed to the need and mandate to further scale across the entire system, she acknowledged that this is difficult and requires time and effort to create and scale the integrated ecosystem and build trusting relationships with evolving community partners around a common goal for the population. But with the data they had collected and the positive results from their evaluation, mental health integration program staff were able to convince the leaders of the health system that it was worth the time and effort to continue investing in the model. She noted the limited mental health resources in the beginning. With no reimbursement, the model was still fee for service in 2000, but it did have an integrated delivery system. Once that was paired with the data supporting the business case, it allowed for shifting resources, building algorithms, and understanding the complexity of patients and the best ways forward.
DISCUSSION
Discussion topics included implementation timelines; definitions, data, and metrics within a system; and how primary care can become more involved in mental health care delivery.
Implementation Timelines
Shannon Harris, associate professor at Virginia Commonwealth University, asked the speakers about timing and how long it typically takes systems to see results after implementing changes. Bitton noted that the IAPT model took about 10 years to be scaled to serve nearly 1 million patients with beneficial results (Clark, 2018). However, because it was built on the chassis of an integrated primary care system, it yielded early good results and proofs of concept that helped carry the momentum to scale. The good news for VA, he added, is that it has a scaffold of a strong integrated system, with strengths in place to build on. But community components will require investment in families, community engagement, and recovery models to reach the outcome goals that people desire, said Bitton. Reiss-Brennan added that implementation never truly ends, especially if you want to build for the long term, prioritize social connectedness, and improve community health and wellness. Intermountain’s implementation journey took 20 years, but that was intentional. Of its five key components that really improved outcomes, she said that the true game changers are leadership and culture. Reiss-Brennan said that leaders need to align their cultural and social mission with infrastructure that connects meaningful data across a system to share the impact of their mission; build the conditions for trusting relationships; bring in new information, pivoting where needed; and reach out to other organizations to build partnerships for success.
Definitions, Data, and Metrics
Within the context of mental health access, Reiss-Brennan said Intermountain already had a definition for high-quality care, based on the Institute of Medicine’s report on health care quality (IOM, 2001). But moving into the community and primary care spaces, she acknowledged, the definition of high quality was more focused on improving efficiencies and lowering costs. Once Intermountain understood the real lived experience that primary care teams had with patients, the organization recognized a mental health and social component of these relationships that needed to be incorporated into the approach to quality. Reciprocity became key, she said, between patients and families and providers and between and across staff members and leadership, because without it, people get burned out, and goals will not be realized.
Harris noted variability in services across the VA system but that in the international models Bitton presented, all had core metrics and training. VA is a positive outlier for the United States in terms of data infrastructure and core training programs, Bitton said, so leaning into standard training around these core models and hiring people on a quicker timeline can help. He argued that VA has advantages in standard data systems and training models but needs to accelerate the bureaucratic hurdles of hiring and find pathways that honor the
diversity between regional communities yet also maintain a core standardization. The medical approach to behavioral and mental health helps to name and diagnose conditions and diseases, Bitton said, but a social and emotional approach is needed for treatment. In recovery from any kind of mental illness, people need three things: people, place, and purpose. To successfully find the third, we need to engage families and communities, he said. This direction will need investment from leadership and may look different across regions, but it is a required step to reach these goals.
Regarding advanced analytics, a participant noted VA’s wealth of amazing data, dashboards, and provider information, but streamlining pathways through advanced analytics is still under development. Reiss-Brennan replied that it was incredibly helpful to have artificial intelligence (AI) and mathematical models added to the complexity models to understand what different patients and clinics were reporting. Intermountain also has a data warehouse with all the electronic health record (EHR) data and longitudinal data mapped to patient identifiers, neighborhoods, and other demographics. Using that, Intermountain built a data measurement scorecard to measure team performance. She acknowledged that a main challenge is that some analysts may be working with certain registries for close to 20 years, but maintaining this wealth of information may be difficult with staffing and organizational changes. Another participant asked about using data in the AI models for things such as diagnosis and prevention. Reiss-Brennan said Intermountain Health has an assessment tool that works with the EHR and identifies all of the relevant data points to pair with other information, such as medication history and number of visits; it predicts the need for screenings for depression and other mental health services or for a social determinant screening.
Improving Mental Health Through Primary Care
Bitton said it is critical for primary care providers to be involved in improving mental health for veterans, and the first thing all good comprehensive primary health care systems do is understand their patient population. Health care often starts inside the box of the clinic, he explained, but VA’s denominator should be every veteran and their family. Many pathways to the clinic exist, such as phone/text programs, national campaigns, or other outreach. He said that we need to do a better job of making it possible to connect with these veterans. Reiss-Brennan suggested normalizing mental health as part of the military experience and going upstream to address these issues earlier in their military careers. Reiss-Brennan said veterans and their families should not have to wait for something to happen and then be expected to deal with it after their service. Rather, upon joining the military, the service branches should educate them and their families about what has happened to others, common experiences, and available resources.
